From owner-freebsd-hackers Mon Dec 20 12:51: 0 1999 Delivered-To: freebsd-hackers@freebsd.org Received: from apollo.backplane.com (apollo.backplane.com [216.240.41.2]) by hub.freebsd.org (Postfix) with ESMTP id E4F2915332 for ; Mon, 20 Dec 1999 12:50:58 -0800 (PST) (envelope-from dillon@apollo.backplane.com) Received: (from dillon@localhost) by apollo.backplane.com (8.9.3/8.9.1) id MAA17151; Mon, 20 Dec 1999 12:50:54 -0800 (PST) (envelope-from dillon) Date: Mon, 20 Dec 1999 12:50:54 -0800 (PST) From: Matthew Dillon Message-Id: <199912202050.MAA17151@apollo.backplane.com> To: Warner Losh Cc: Wes Peters , freebsd-hackers@FreeBSD.ORG Subject: Re: Cool little 100BaseTX switch - they're coming down in price References: <199912202024.MAA17008@apollo.backplane.com> <199912190410.UAA01049@apollo.backplane.com> <385C60FC.7613CB55@bellatlantic.net> <19991218225758.A23729@futuresouth.com> <199912190556.AAA08484@whizzo.transsys.com> <199912191943.LAA06826@apollo.backplane.com> <385D47D3.FCEE9EAB@softweyr.com> <199912192127.NAA09156@apollo.backplane.com> <385DDE7A.1A0ED466@softweyr.com> <199912201949.LAA16719@apollo.backplane.com> <199912202034.NAA18342@harmony.village.org> Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG :Yes. The nice thing about modern swithcing power supplies are that if :you DO hear any vibration, you know you have big problems and are :courting disaster. At least that's my experience in homebrewing a 12V :-> 16V converter for my Sony VAIO 505TS. The original magnetics I :chose easily overloaded, producing a noticable hum. I didn't think :much of this at the time, but about 3 hours into my test the coil :started smoking a little, and then the cascade failure from that :caused an overload on the regulator ic I was using and after the pop :no more hum, but no more +16V either :-( And I was only running at :20 kHz.... : :Warner Also, step-up switching supplies are a lot more finicky then step-down supplies. We usually use a flyback configuration for step-up but the input voltage range tends to be more constrained and the efficiency is lower so I try to avoid step-up whenever possible. I used a step-up (flyback) in one of my telemetry boards - incredible sucker, it took 12V in and produced 24V at 3A. I made one mistake in the design and that was to use too low a wattage on the 0.05 ohm current limit resistors (the last thing you'd expect to overheat). My personal preference is to use an *unregulated* DC wall plug (which is essentially just an AC transformer, rectifier, and big caps) that produces somewhere on the order of 15-25V and then put a step-down switching supply on the motherboard. This way we don't really care what the DC power going into the device is, or how noisy it is, just as long as it's at least 1.5V more then our regulator output. It lets us use virtually any wall plug, or a solar panel, battery, or DC power without making a single change to the design. The standard national switching supply chips can handle DC inputs up to, what, 40V? I think they have an industrial (read: for a car) version that can handle even more. They are efficient enough that even worse case inputs have low enough heat dissipation. I usually use the 1A parts though one recent design uses their 5A part. If we needed more power we used to use TO-220 packaged PFETs for the switching circuit. Now days we use SMT duel-PFETs which are actually rated for higher current in plastic then the old PFETs were in ceramic. But for low-current (< 3A) applications the national parts can handle the switching natively, which I really really like. -Matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message